PROJECT SUMMARY Home health care (HHC), the most rapidly growing health care sector in the U.S., plays a significant role in post-acute care. Over 4.7 million patients received HHC in 2011, and over half had an inpatient stay immediately before HHC admission. Infection is prevalent in home care; in previous NIH-funded work, we found that three of the top five reasons for hospitalization among HHC patients were related to infections. In fact, infection control was listed by The Joint Commission as one of its national patient safety goals for HHC. This indicates the importance of focusing on infection control and prevention in HHC in national efforts to reduce re-hospitalization. Critical steps before implementation of any infection control intervention include identifying patients who are potentially at high risk for infections, and characterizing current infection control practices in HHC. Clinical decision making is often limited by a lack of systematic assessment of the complex needs of patients and their environments; specifically in the HHC setting, little is known about staff infection control practices and how high risk patients are identified. To address these challenges, the aims of this project are to 1) develop and test a predictive risk model to identify patients at high risk of infection in HHC setting, and 2) assess current infection control practices in the HHC setting and explore how predictive risk modeling can be utilized in clinical practice in HHC setting. The proposed project will be conducted in the Visiting Nurse Service of New York (VNSNY), the largest not-for-profit HHC agency in the U.S. For Aim 1, a predictive risk model will be developed in 70% of the sample and then validated in the remaining 30% of the sample. The predictive risk model decision will also be validated with HHC infection control specialists? clinical judgment. For Aim 2, mixed methods including observation, survey, semi-structured interviews and focus groups will be used to provide a comprehensive understanding of current infection control practices and potential barriers. Our research team has substantial expertise in nursing, decision analysis, modeling, epidemiology, mixed methods, infectious disease, and infection control and access to the rich statistical resources from Columbia University Medical Center and the Center for Home Care Policy and Research in the VNSNY, therefore is ideally qualified to conduct this study. The study findings are the critical first step toward future development of an efficient risk model guided infection control intervention in the HHC setting.